Appearance
A genetic edition encodes how a text was written — the sequence of additions, deletions, and substitutions an author made — rather than presenting a single finished reading. In TEI you build it from a small core of elements (<add>, <del>, <subst>) anchored to the document surface (<sourceDoc>, <surface>, <zone>) and sequenced by writing stages declared in the header. Start with one heavily revised paragraph, link it to a facsimile, and get the stages right before scaling up; genetic encoding rewards patience and punishes haste.
What is a genetic edition in plain terms?
Imagine an author's draft covered in crossings-out and insertions between the lines. A genetic edition is the scholarly attempt to answer "in what order did all that happen, and what did the page say at each step?" It treats the manuscript not as a container for a final text but as a record of a process. This is why it matters to literary scholars: the deletions and second thoughts are the evidence of how a work came to be, and a clean reading text throws that evidence away.
How is it different from other editions?
| Edition | Question it answers | Primary witnesses |
|---|---|---|
| Documentary | What does this page say? | One |
| Critical | What is the authoritative text? | Many |
| Genetic | How did the text come to be? | One author's drafts |
A genetic edition usually focuses on authorial manuscripts where you can see revision happening. It overlaps with documentary work (you care about the surface) but adds a temporal axis the others lack.
Which TEI elements do beginners actually need?
You do not need the whole genetic module on day one. Start with four elements:
<add>— text the author inserted.<del>— text the author struck out.<subst>— a deletion-and-replacement pair, grouping a<del>and an<add>.<surface>/<zone>— where on the page the change sits.
Here is a single revised line, encoded minimally:
xml
<line>The sky was
<subst>
<del>blue</del>
<add place="above">grey</add>
</subst>
that morning.</line>This says the author wrote "blue", struck it, and wrote "grey" above the line. Already you have captured a revision, not just a result.
How do I show the order of revisions?
The hard part of genetics is sequence. TEI lets you declare writing stages in the header and point changes at them:
xml
<!-- in the header -->
<creation>
<listChange ordered="true">
<change xml:id="s1">First draft, ink</change>
<change xml:id="s2">Revision, pencil</change>
</listChange>
</creation>
<!-- in the text -->
<add change="#s2" place="margin">in the cold light</add>Now a processing tool can reconstruct the page as it stood after stage one, then after stage two. That staged replay is the payoff of genetic encoding.
Why must I link to the facsimile?
Genetic claims are arguments about physical evidence — which insertion came first, whether a deletion is the author's or a later hand. None of that is checkable without the image. Anchor your transcription to an IIIF facsimile so a reader can verify every claim:
xml
<surface facs="https://images.example.org/iiif/draft1/canvas/p3">
<zone ulx="200" uly="450" lrx="1900" lry="620">
<line>...</line>
</zone>
</surface>Most published genetic editions show the manuscript and the staged transcription side by side precisely so the evidence is one glance away.
What is a sensible first project?
Do not start with a whole notebook. Choose one paragraph with three or four clear revisions, photograph or source its IIIF canvas, and encode it through all the steps above: surface, zones, <subst>, and at least two declared stages. Build a tiny stylesheet that renders stage one and stage two separately. When that small example works end to end, you understand the model and can scale to a folio, then a quire. Trying to encode everything at once is how beginners abandon genetic editions.
Key Takeaways
- A genetic edition encodes the writing process, not just the final text.
- The core beginner elements are
<add>,<del>,<subst>, plus surface and zone. - Declare writing stages with
<change>and point edits at them via@change. - It differs from critical editions: one author's drafts, asking how the text formed.
- Always link to IIIF facsimiles; genetic claims need verifiable images.
- Start with one revised paragraph and a two-stage replay before scaling.
- Genetic encoding is demanding because it adds a temporal axis to transcription.
Frequently Asked Questions
What is a genetic edition?
A genetic edition reconstructs how a text came to be — the order of additions, deletions, and revisions an author made. Instead of presenting a final text, it shows the writing process as a sequence of stages, which is invaluable for studying authorial intention and creativity.
Which TEI elements do I need for genetic encoding?
The core set is <add>, <del>, <subst>, <restore>, and <retrace>, plus <addSpan>/<delSpan> for long stretches. For the document surface you use <sourceDoc>, <surface>, and <zone>, and you sequence changes with <change> and the genetic @change pointers.
What is the difference between a genetic and a critical edition?
A critical edition reconstructs one authoritative text from many witnesses. A genetic edition reconstructs the stages of a single text's creation within one or a few authorial witnesses. Critical asks 'what is the text?'; genetic asks 'how did it become the text?'
How do I show the order of revisions?
Use the TEI genetic apparatus: declare writing stages with <change> in the header, then point each <add> or <del> at a stage with the @change attribute. A processing step can then replay the document stage by stage.
Do I need facsimiles for a genetic edition?
Strongly recommended. Genetic claims about insertions and deletions are only verifiable against the image, so link your encoding to IIIF facsimiles. Many genetic editions present the manuscript image and the staged transcription side by side.
Is genetic encoding harder than other editions?
Yes, it is among the most demanding because you encode not just text but the temporal sequence and spatial position of every change. Start small — one heavily revised paragraph — before attempting a whole manuscript.